🚀 go-pugleaf

In article <wZfVc0200...@andrew.cmu.edu> sa...@andrew.cmu.edu (Sudheer Apte) writes:
>                        ...I'd just like to make a little correction
>to the "rm *" discussion.  Somebody pointed out the common mistake of
>not SHIFT'ing in time when typing "rm *.o", getting "rm *>o", and
>claimed that you get just one empty file named "o".  Actually, the
>shell creates the file "o" before calling rm, so the command unlinks
>that, too;  you get nothing at all in your directory.  (Not that you'd
>be worrying about this minor fact when all your sources are gone,
>but...)
	
	I'm not trying to pick nits, but...

	No, you're wrong.  Well, you're right, but you're wrong.

	Yes, the shell creates the 'o' file before calling rm, but before
	doing that, it also expands the *, giving all the filenames in the
	directory (sans o) and passes them to rm.  

	I assume that this is the standard order for most *nix shells. (?)
	It's at least true for ksh running under our SysVr3

						Derek
-- 
.sig file closed for repairs--No thru traffic

[Followups redirected to comp.unix]

In article <wZfVc0200...@andrew.cmu.edu> sa...@andrew.cmu.edu (Sudheer Apte) writes:
>not SHIFT'ing in time when typing "rm *.o", getting "rm *>o", and
>claimed that you get just one empty file named "o".  Actually, the
>shell creates the file "o" before calling rm, so the command unlinks
>that, too;  you get nothing at all in your directory.

Ah, but the shell also expands '*' before calling rm!  The question is,
which does it do first?

[Exercise is left to the reader.]
[Hint:  "the shell" is misleading.]
-- 
cha...@C3.COM	{decuac.dec.com,cucstud}!c3pe!charles	ex::!echo Boo:

I did an equal thing when I was in high school (Crete, Nebraska,
but that is another story). 

On Apple ][ the disk I/O in Basic was done via printing
utilities, chr$(13)chr$(4) was the code to start a DOS command. I
am skipping a lot of details, but one of the commands was new,
which clears the Basic workspace.

Anyway, we were supposed to program something simple in math
class. My program calculated the sides and angles of a triangle.
But besides showing the program we were supposed to give the code
to teacher, too. So I had put several lines of comments in the
beginning of program and then I went to the monitor (like
debugger) mode and changed some blanks in those comments to
carraige returns, backspaces and ctrl-D:s. 

So I turned in the source code and the teacher went happily home
with it. Next day he complained to the class that my program
seems to calculate the triangles all right, but he can't
undestand how it does it, since all he can get with the list
command is a blank screen with my copyright message. I had used
backspaces and blanks to cover the line numbers and placed a CR,
ctrl-D, NEW, CR on the last line, which cleared the buffer. 

Of course this is easy to crack by printing the code on paper,
slowing printing to screen or with a debugger. But at those days
it was a lot of fun. 

Little things like these work still. You can place a ctrl-Z to
the start of MS-DOS file, this stops the type-command. Or having
a line containing only word DISCONNECT in it, does interesting
things with some communications programs. 
--
Timo Kiravuo
Helsinki University of Technology, Computer Center
Tel: work 451 4328, home 676 076, Fax 466 251;  Intl. add +358 0
kir...@hut.fi  sorvi::kiravuo  kiravuo...@uunet.uu.net

In article <wZfVc0200...@andrew.cmu.edu> sa...@andrew.cmu.edu (Sudheer Apte) writes:
   Before I relate my story,  I'd just like to make a little correction
   to the "rm *" discussion.  Somebody pointed out the common mistake of
   not SHIFT'ing in time when typing "rm *.o", getting "rm *>o", and
   claimed that you get just one empty file named "o".  Actually, the
   shell creates the file "o" before calling rm, so the command unlinks
   that, too;  you get nothing at all in your directory.  (Not that you'd
   be worrying about this minor fact when all your sources are gone,
   but...)


I guess this depends on your shell.  sh does wildcard expansion
before it creates redirection files.  csh creates the file, then does
wildcard expansion.  I personally think that csh does it wrong.


	   > LIST
	   Pirate! BLECHHH!

	   > _

   Pretty frustrating, and the code is still lying around at the BHU
   computer center. I wonder if anyone knows how it works....

Send it to a printer.  People were always pulling stunts like this on
the old Commodore PETs.  The printers would just produce a graphics
symbol for non-printable characters and the whole listing would be
generated intact.  Your printer may differ.


--

Andrew Thomas
and...@watsnew.waterloo.edu	Systems Design Eng.	University of Waterloo
"If a million people do a stupid thing, it's still a stupid thing." - Opus

In article <wZfVc0200...@andrew.cmu.edu>, sa...@andrew.cmu.edu (Sudheer Apte) writes:
> The BASIC not only allowed any characters inside comments, but added
> its own control codes to the basic screen driver routines (the BASIC
> was resident in a ROM and loaded up at boot time along with the OS,
> so the machine was always in BASIC expr-eval mode), making it possible
> to save an ascii file full of control characters and simply EXEC'ing
> it from BASIC... What we did was: wrote a nice long program to do
> graphics on the screen, and exit with a taunt to ``try reading the
> code to discover how this is done!''. Here's the program:

I used to have fun poking a Commodore PET basic program so that the
only lines that LISTed were ones that were not the target of a GOTO.
These had to be contained in the linked list because the interpreter had
to be able to search for them; but sequentially executed statements could
be left out of the list because it just kept going and didn't search for
the next line.

Unfortunately, I soon discovered that you couldn't SAVE a program this way.
When you LOADed it back, the links were restored. :(

===============================================================================
Roger Ivie

35 S 300 W
Logan, Ut.  84321
(801) 752-8633
===============================================================================

In article <KIRAVUO.90...@kampi.hut.fi> kir...@kampi.hut.fi (Timo Kiravuo) writes:
o On Apple ][ the disk I/O in Basic was done via printing
o utilities, chr$(13)chr$(4) was the code to start a DOS command. I
o am skipping a lot of details, but one of the commands was new,
....

There was something else which was undocumented on these machines (and
which I used about 11 years ago)...  If you inserted a magic series of,
I believe, 5 bytes (I don't remember what they were anymore) between
the line number and the tokens for the BASIC program itself, you would
selectively make the line of code invisible.  A LIST would give a list
of the line numbers, but not the code, though the program itself would
work without a hitch.


-charles
-- 
Charles Daffinger  >Take me to the river, Drop me in the water<  (812) 339-7354
cd...@iuvax.cs.indiana.edu             			 cd...@ed.ecn.purdue.edu
Home of the Whitewater mailing list:    whitewate...@iuvax.cs.indiana.edu

In article <wZfVc0200...@andrew.cmu.edu> sa...@andrew.cmu.edu (Sudheer Apte) writes:
>Now for the lore:
>
>This happened back in BHU Varanasi, India. The university computer
>center had just been given a set of BBC Microcomputers (for
>educational computing in schools) for training schoolteachers on.
>These BBC Micros were beauts: 8-bit 6502 machines, they came with
>their own color monitors with 16 colors, 5 1/4 " floppy drives, and a
>native Pascal-like BASIC. They had lots of neat options, too-- D/A


Talking of hidden messages, on a BBC micro, if you typed :

          RENUMBER 10,2345

it came back with the message :

          silly

This was the only thing you could do that would give this message.

And, of course, the manual claimed that anything you typed in
at the keyboard could not possibly damage the hardware. This was
not true. The machine had a cassette port with remote control
to switch the motor on and off which utilised a reed relay,
so if one was to type the short program

          10 *MOTOR1:*MOTOR0:GOTO10

nasty things would ensue.
 


-- 

To be is to do - Socrates          |Jim Leitch, Dept of CS, University of
  To do is to be - Sartre          |Strathclyde, Glagow, European City of
    Do be do be do be do - Sinatra |Culture, Scotland.................
standard disclaimer : "It wasn't me, anyway I was on holiday at the time" <8'} 

In article <14...@baird.cs.strath.ac.uk> jle...@cs.strath.ac.uk writes:
>
>And, of course, the manual claimed that anything you typed in
>at the keyboard could not possibly damage the hardware. This was
>not true. The machine had a cassette port with remote control
>to switch the motor on and off which utilised a reed relay,
>so if one was to type the short program
>
>          10 *MOTOR1:*MOTOR0:GOTO10
>
>nasty things would ensue.
> 

Also on the beeb, removing the system ROM, and looking at it in an EPROM
programmer, say, revealed a list of credits for people involved in the 
construction etc. You could not read this insitu as it was under memory mapped 
devices, and the ROM was disabled when the devices were active.


jon

-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
=                                                                           =
-                yet to get a .signature file                               -
=                                                                           =
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

jle...@cs.strath.ac.uk (James T Leitch CS88) writes:

>Talking of hidden messages, on a BBC micro, if you typed :
>          RENUMBER 10,2345
>it came back with the message :
>          silly

>This was the only thing you could do that would give this message.

   Yes. And Amcom (who produced a rather natty little Network Operating 
System for the Beeb that knocked the spots off Acorn's Econet) put in an early
version of their E-Net NFS workstation ROM a message that appeared if you tried
to do something to someone else' files:

            naughty.

   May I also briefly note that if you put a DFS 0.90 ROM the wrong way round
in a Beeb Micro then turn the power on, it is possible to get a loud pop as the
quartz window out of the top of the EPROM flies vertically upwards. I've seen
it done, it's ever so pretty (but it does leave rather a pong of burning chip
in the room ...) .

	Dave C, UEA, Norwich.

In article <16...@cc.usu.edu> SL...@cc.usu.edu (Roger Ivie) writes:
>I used to have fun poking a Commodore PET basic program so that the
>only lines that LISTed were ones that were not the target of a GOTO.
>These had to be contained in the linked list because the interpreter had
>to be able to search for them; but sequentially executed statements could
>be left out of the list because it just kept going and didn't search for
>the next line.
>
>Unfortunately, I soon discovered that you couldn't SAVE a program this way.
>When you LOADed it back, the links were restored. :(

I did this with Data General DG BASIC on an Eclipse running AOS 4.xx.
Funstuff. The trick then was that it used 16 bits to represent the line number,
but only let you enter numbers up to 9999. I sat up one night figuring
out the tokenization method, and wrote a program that would read a
saved BASIC file, alter all the line numbers, (GOTOs and GOSUBs as well,
of course) and voom - the lines vanished. Unfortunately, a simple
RENUMBER command would correct the situation. But, there was a way around
that as well - made one of the first of the illegally numbered lines a
GOTO xxxx where xxxx didn't exist. Then renumber would just choke and give
up. I guess I'd just heard of synthetic programming on the HP-41 back
then, so I played with everything available, in great detail. The same
trick in Applesoft would give you lines that could be listed but couldn't
be deleted. Great for copyright notices.

I was pretty fond of the DG BASIC trick, tho. Wrote an entire Adventure
game, using overlays & all kinds of tricks to fit into 32K core. The
listing only had 10 lines of REMs, the code was nowhere to be seen. I
was well pleased with myself.  }-)  One of the other neat tricks was to
set the length of a line to be zero, then the interpreter's program counter
never advanced, and you'd get an infinite loop. You could set the
high bit too, getting negative offsets. Really wonderfully fast GOTOs,
but you had to count bytes pretty carefully, and it made your listings
pretty Strange and Wonderful. (Keep a hand on the Escape key...)
--
 -=- PrayerMail: Send 100Mbits to holyghost@father.son[127.0.0.1]
 and You Too can have a Personal Electronic Relationship with God!

>>From ukc!sys.uea!cmp8118 Wed Jan 17 01:20:50 GMT 1990

>   May I also briefly note that if you put a DFS 0.90 ROM the wrong way round
>in a Beeb Micro then turn the power on, it is possible to get a loud pop as the
>quartz window out of the top of the EPROM flies vertically upwards. I've seen
>it done, it's ever so pretty (but it does leave rather a pong of burning chip
>in the room ...) .
>

   The BBC's are very good at blowing up, when the first batch of the BBC
Master came out they had a battery backed set of options. At the last minute
the Nicad rechargables where replaced by normal dry cells, leaving the charging
circuit behind. One simple rom call and an hour or so later... BANG!!!! no more
BBC. This could be used to make one hell of a virus!

Dave H.